Recursive tuning algorithm for assist controller of a trolley crane system

The role of a human operator in machine control varies with the level of automation. In applications where the human operator is directly in control, the most significant performance variations are due to the variations in the human performance. Human control is very dexterous and is especially advantageous in complex multi-objective task execution. However, the human performance variations might compromise the safety and accuracy of the system. To compensate the insufficient human performance, parallel human adaptive assist control is developed. The assist control is tuned based on the human operator model identified during normal task execution. Moreover, a new recursive tuning method for the human adaptive assist control is described.